Stuart Henderson

FWIW I haven't seen any icinga crashes on OpenBSD recently. As far as reports go, this is quite sparse, no details of config, no debug logs, minimal detail in the...

Ah I think this explains something I had noticed occasionally with director but hadn't figured out how to reproduce reliably. I can't say 100% for sure, but I think this...

I see the same segfault with numpy 1.23.2 on OpenBSD/i386, including with Python 3.10. (I'm trying to prepare the OpenBSD ports tree to move to Python 3.10 by default, and...

Similar segv/backtrace if I build with GCC instead of Clang.

Aha, after some bisecting it seems it was introduced in 8245b392a344a (the big "ENH: enable multi-platform SIMD compiler optimizations" diff). Moving back to 1.23.2 I've tested with "--cpu-dispatch min", which...

@didickman right. As well as amd64, I also have positive test reports on arm64 and riscv64 now. The machine I'm doing i386 tests on is EPYC 3251, so the cpu...

It is, I'm running a 32-bit-only OS version on it for package builds/tests. But if it's just an identification problem, wouldn't --cpu-dispatch min avoid that? (that looks like it only...

I'm hitting an issue on 32-bit builds with the libutil/cxx/file_util.cxx changes in 3.4, ``` /pobj/rspamd-3.4/rspamd-3.4/src/libutil/cxx/file_util.cxx:209:49: error: non-constant-expression cannot be narrowed from type 'long long' to 'std::size_t' (aka 'unsigned long') in...

Correct, it persists after restarting lldpd. (Bringing the interface down and back up clears it, at least for a while). IFF_RUNNING doesn't indicate link state; in if.h the comment says...

Great, 922b3fb68eebf works, many thanks Vincent!